Обобщая- авторы пришли к выводу, что введение такой меры, как ношение маски в общественном месте и пр, на уровне штата (или страны) приводит к тому, что в популяции на 13% сокращается заболеваемость ковидом, на 7% госпитализация, и, в итоге, смертность от ковида- на 20%.
We therefore first create a dataset of state-level mask mandate start and end dates by manually reading each state government’s memos
We then build on recent work using event study designs to estimate treatment effects in the context of theCOVID-19 pandemic: the effect of NPIs (e.g. business closures and stay-at-home orders) on the volume ofonline searches related to unemployment , on social distancing in the US , on the airline industry ,on stock markets worldwide [41–44], and on COVID-19 cases at the county level . For background, theeffect of the introduction of new policies (here, mask mandates) has been studied with a variety of techniques such as difference-in-differences , event studies  and regression discontinuity .
An event study design allows us to estimate the treatment effect associated with mask mandates onCOVID-19 outcomes on each day following the introduction of the mandate relative to the day prior to its introduction. Our geographical unit of analysis are states to minimize peer-effects of neighboring counties’ mandates due to the underlying interdependence between county mobility patterns , as well as the factthat people living in one county often have to travel to a different county to get medical care, resulting in inconsistent accounting of COVID-19 health outcomes at the county level .
Although previous work has investigated the effect of mask mandates on COVID-19 outcomes [7–10], our results are novel because our event study specification simultaneously accounts for the following: all 50 USstates and the District of Columbia, longer timescales (up to 50 days after the introduction of a mandate,and for the time period between February 1 and September 27, 2020§), three COVID-19 outcomes: daily new confirmed daily cases, daily new confirmed deaths, and the proportion of daily new hospitalization admissions due to COVID-19.To support our result of the effect of mask mandates on COVID-19 outcomes, we investigate the associated increase in mask adherence (i.e. the percentage of people who wear masks in public) following a mask mandate for the four states – Hawaii, Iowa, North Dakota and New Hampshire – for which adherence data  is available before and after the introduction of the mandate, as detailed in section 1.2.
We estimate the effect of mask mandates on three prominent COVID-19 outcomes: 1) number of newconfirmed cases, 2) proportion of newly-admitted COVID-19 related hospital admissions (i.e. the number of patients admitted to treat COVID-19 symptoms relative to the total number of admitted patients), and 3)number of deaths. We ensure to only use new cases, hospital admissions and deaths each day, as opposed to cumulative numbers.
For daily new cases, we observe a delay of about 11 days after the mandates are introduced for thetreatment effect to start increasing (i.e. going negative). The treatment effect continues to increase (i.e. itsmagnitude becomes more negative) reaching -0.45 standard deviations (95% confidence interval [-0.68,-0.26])50 days after the introduction of mask mandates (adjustedR2is 0.431 andp <0.001). The magnitude ofthe associated treatment effect is large, corresponding to 3.24 cases per 100,000 people or 13% of the highestrecorded number of new cases per 100,000 people during our observation period.
The delay for mask mandates to have an associated treatment effect on COVID-19 related deaths is 19days after the start of a mandate. This 8-day time lag between the decrease in the number of cases and thenumber of deaths is in line with the temporal pathogenesis characteristics of the virus as the interquartilerange of the time between symptom onset and death is 8-26 days [53–60]. After 19 days, the treatment effectcontinues to increase (adjustedR2is 0.304 andp <0.001) reaching -0.80 [-1.02,-0.59] standard deviationsafter 50 days which corresponds to 0.19 deaths per 100K, or about 20% of the highest recorded number ofdaily COVID-19 related deaths observed during our observational period
Finally, we also estimate the effect of mask mandates on the only type of hospitalization data availableat the state level for all states since the beginning of the pandemic (here, as early as February 1, 2020): theproportion of daily hospitalization admissions‡ due to COVID-19. It is important to note that we onlyobserved theproportionof hospitalizations, which is the the number of new hospitalizations due to COVID-19 relative to the number of total admissions
Similarly to cases and deaths, we find that the introduction of mask mandates leads to a statistically significant decrease in the proportion of COVID-19 related hospitalization admissions after 8 days andit continues to decrease thereafter (adjustedR2is 0.511 andp <0.001). 50 days later, the proportion of OVID-19 related hospitalization admissions decreased by -2.47 [-3.39,-1.54] percentage points (not standarddeviations, since this outcome is an already normalized percentage). For reference, this corresponds to a 7%decrease compared to the highest recorded proportion (34%) of COVID-19 related hospitalization admissions during our observation period.